Sleep onset is one of the most robust and dramatic changes that the human brain normally undergoes. However, the inability to initiate sleep (insomnia) or appropriately restrain it (excessive daytime sleepiness, narcolepsy) adversely affects millions of people each year, resulting in large personal as well as societal costs. Animal research from the last 10 years has produced remarkable advances in our knowledge of the local cellular and neurochemical mechanisms that regulate sleep onset. Discrete hypothalamic nuclei have been implicated in the active production of sleep through the selective inhibition of key behavioral state control centers in the brainstem. The physiological study of human sleep has largely been limited to EEG techniques, which suffers from poor spatial resolution. Work using SPECT and, more recently, PET has revealed complex dissociated patterns of functional neuroanatomy associated with the sleep stages of NREM and REM, findings that past electrophysiological studies could only hint at. Despite these findings, there exists a clear knowledge gap regarding the neural mechanisms responsible for the sleep onset transition in the human brain, at a whole systems level. Using the superior spatial and temporal resolution of functional magnetic resonance imaging (fMRI), combined with simultaneous electroencephalography (EEG), we have recently demonstrated the ability to record changes in regional blood oxygenation levels in both subcortical and cortical regions across sleep onset in humans. The data provide evidence for dissociated shifts in activation levels that are tightly linked to the time of sleep onset in brainstem, thalamic, and hypothalamic nuclei, as well as shifts in cortical activation. The studies proposed here are designed to provide sufficient pilot data to establish the value of this research paradigm, with the larger goal of developing specific hypotheses from these pilot studies that can be tested using fMRI combined with sleep polysomnography (PSG) and, through collaborative ventures, in animal models as well.